Description

The UP COPC Research Unit’s modelling and baseline analysis tool has been designed to run a baseline analysis for the whole country using its Integrated Health System Planning (iHSP) toolkit. This baseline (level 1) analysis creates catchment areas using StatsSA 2011 Census small area population data (updated to 2018) around all clinics and community health centres. The output from this analysis includes: provincial reports showing facility level demography (population profile and size, income and deprivation index) and geography of each catchment in tables and on a user accessible GIS, mapped community profiles, and this information is linked to districts and sub districts.

Once the baseline analysis is complete, the UP COPC Research Unit update and expand baseline analysis for target community using iHSP. The population update include GTI^[1] population dataset and map-making adjustments; localised burden of disease factors; and refined catchment area boundaries (local geography). The output from this analysis include: adjusted catchment area analysis report with profile of target community (demography, income, deprivation, geography – datasets and mapped output); projected workload for and numbers of CHWs required; required enabling staff and equipment schedules; costing schedules; and resource availability mapping.

Critical success factors for the required improvement in service delivery and outcomes are:

• Making the home the start and end point of services thereby enabling the shift from reactive treatment to proactive prevention.
• Ensuring service, information and relationship continuity of care through managed care coordination, up and down levels of service as well as across disciplines and sectors.

The baseline needs analysis must be based on the resources required in the contracting unit to ensure those conditions are met.

Rationale and guiding principles

1. Core objectives of the baseline modelling process:

1. Determine and profile the communities that make up health catchment areas (HCA) which, when aggregated, describe the geographic population to be supported by a contracting unit (service purchaser) and serviced by contracted providers.
2. Attach communities in the HCA to their closest PHC provider unit (initially public sector clinics and community health centres).
3. Identify greatest vulnerability within segments of the population to ensure critical impact and to enable incremental roll out of NHI services.
4. Provide a baseline assessment of the required first tier of healthcare services (CHWs and team leaders) attached to each provider unit in the CU.
5. Provide a baseline assessment of the numbers of clinicians (nurses, clinical associates, doctors and specialists, as well as allied health professionals) required to deliver coordinated care.
6. Provide a baseline assessment of the suitability, condition and size of PHC facilities to comply with a minimum acceptable standard for PHC facilities.
7. Estimate the cost of delivery of the COPC service, including linkage to care and care coordination within the referral chain for complex care, for NHI budgeting and allocation requirements.

 

 

 

2. Analytical methodology

2.1 StatsSA census data and mid-year estimates of population growth are considered robust at District level only.

They are not robust below District level for two reasons:

1. Mid-year estimates do not provide insight into spatial distribution within the district. This makes linear projections of census populations below district level often misleading. New RDP housing, informal settlements[2] or back yard densification can change catchment populations for a provider unit by 100% or more.
2. Previously unpopulated areas that have now become populated are not spatially distinguishable[3] in Stats SA structures. New coding is required to identify growth nodes and allocation to an appropriate provider unit.

2.2 There is a need therefore for detailed analysis (visualisation over time) of the CU to identify significant growth areas or densification within or around communities. Once identified, further analysis is required to estimate numbers of households and population.

2.3 StatsSA data provides total population broken down by age and gender, and households broken down by income strata. Since household size is linked to income levels it is necessary to establish household numbers by income quintile or income group and then estimate the numbers of people in each quintile or group by weighting household size.

2.4 Communities need to be described with sufficient granularity to be able to create a many-to-one relationship with provider units to which they are attached.[4]

1. StatsSA small area layer (SAL) data set is the highest granularity published and should be used. The enumeration area dataset is only available commercially and provides little extra detail.
2. Communities should be attached to their geographically closest provider unit[5] using a “nearest neighbour” analysis of the SAL on a GIS system.

2.5 Catchment populations must be adjustable to make allowance for physical geography (mountains and rivers) or infrastructure (highways, tunnels or road networks) in order to identify the most accessible service providers.[6]

1. There is a need to visualize the CU catchment area on a satellite image or hybrid^[7] base map.
2.  These make it possible to reallocate nearest neighbour SAL to alternative, most-accessible facilities within the system.

2.6 Heat maps of population numbers and poverty are necessary for initial identification and comparison of potential areas for development of a CU.

2.7 Service structure management and funding are based on health sub districts (SD), so aggregations of HCA and CU should be coterminous with SD boundaries.

2.8 Wards do not provide an appropriate framework for HCA and CU for three reasons:

1. They are lcal political governance constructs which change from time to time for electoral purposes.
2. Many small communities are divided across several wards, which can disrupt the coordination of care..^[8]
3. There are also no structures within the ward system to manage health teams

2.9 Clinical staffing across the CU should be based on workload using the Workload Indicators of Staffing Need (WISN) principle of service contact time available, individual contact time and allowing for a turnover interval between contacts. Initial recommendations for staff profiling are:

1. Community based teams should comprise between 6 and 12 CHW under a single team leader. Each team should have a facility based CHW providing linkage to care.
2. Care coordination between the referral facility and community services will be provided by one clinical associate for 3 teams and one family physician for 5 clinical associates.
3. Ten percent of PHC headcount will require referral to a doctor (MO) and 20% of those (2% of total) will require onward referral to specialists for more complex care.

2.10 Funding and resource management must be based on risk adjusted capitation, not capitation alone, so workload must be adjusted for poverty and burden of disease[9].

1. The profile of CHW visits must be based on the expected CHW contribution to the risk adjusted recommended profiles of patient contacts described in PHC reengineering[10] and adjusted for high medium and low risk groups within the conditions covered /described.
2. Professional nurse workload must be based on the risk adjusted recommended profiles of patient contacts described in PHC engineering⁹.
3. Demand should be nuanced by local (SD)[11] prevalence, incidence and mortality rates.
4. Predicted workloads should be able to be based on current coverage rates and adjusted to achieve target coverage of 95% by 2025.

2.11 Assessment of PHC facilities should be considered in 3 dimensions: Suitability, Condition and Capacity.

1. Suitability: patient flow, light, ventilation, appropriate spaces. Measured on a scale of 1 (unfit for use) to 5 (entirely functional).
2. Condition: basic fabric, fittings and services. Measured on a scale of 1 (unfit for occupation) to 5 (as new).
3. Capacity: Alignment with demand based on planning units – adequate numbers of consulting rooms and spaces supporting them.

2.12 Data collection in the course of service delivery as well as through specific research is necessary

1. To support continuous, iterative refinement and recalibration of the planning tool(s).
2. To enhance efficiencies in resource allocation and performance;
3. To support scale-up.

 

 

The people involved

Click on the name to view the biosketch.

• Rod Bennett
• Geoff Abbott

 

Back to Our Capabilities